Transcriptome Profiling Reveals Exposure to Predicted End-Of
Total Page：16
File Type：pdf, Size：1020Kb
www.nature.com/scientificreports OPEN Transcriptome profling reveals exposure to predicted end-of- century ocean acidifcation as a stealth stressor for Atlantic cod larvae F. H. Mittermayer1*, M. H. Stiasny 1,2, C. Clemmesen 1, T. Bayer 1, V. Puvanendran3, M. Chierici4, S. Jentoft5 & T. B. H. Reusch1 Ocean acidifcation (OA), a direct consequence of increasing atmospheric CO2 concentration dissolving in ocean waters, is impacting many fsh species. Little is known about the molecular mechanisms underlying the observed physiological impacts in fsh. We used RNAseq to characterize the transcriptome of 3 diferent larval stages of Atlantic cod (Gadus morhua) exposed to simulated OA at levels (1179 µatm CO2) representing end-of-century predictions compared to controls (503 µatm CO2), which were shown to induce tissue damage and elevated mortality in G. morhua. Only few genes were diferentially expressed in 6 and 13 days-post-hatching (dph) (3 and 16 genes, respectively), during a period when maximal mortality as a response to elevated pCO2 occurred. At 36 dph, 1413 genes were diferentially expressed, most likely caused by developmental asynchrony between the treatment groups, with individuals under OA growing faster. A target gene analysis revealed only few genes of the universal and well-defned cellular stress response to be diferentially expressed. We thus suggest that predicted ocean acidifcation levels constitute a “stealth stress” for early Atlantic cod larvae, with a rapid breakdown of cellular homeostasis leading to organismal death that was missed even with an 8-fold replication implemented in this study. Global change, caused by diverse anthropogenic activities, is the defning characteristic of the Anthropocene1 and has already started to afect marine ecosystems (reviewed by Doney et al.2). One of the major frst order efects with direct causal linkage to human activity is ocean acidifcation (OA). It is caused by the uptake of rising atmospheric CO2 concentrations, from fossil fuel burning and altered land use, by ocean waters. Te increases in 3 ocean pCO2 and the resulting decreased ocean pH as well as lowered carbonate saturation state have been shown to impact marine ecosystems3–5, particularly calcifying species including foundation species such as corals6. For marine fsh species, results are more complex, i.e. change in thermal window range for diferent life stages7. Most 8–10 adult fsh seem to be less susceptible to increased pCO2 compared to early life stages . Due to their high capabil- 11 ity of acid-base regulation some adult fsh can tolerate pCO2 levels of >8000 µatm , far exceeding the projections 12 of ocean acidifcation for the near future . In contrast, exposure of early life stages to increased pCO2 has been shown to induce severe efects on their performance such as decreased hatching success10, increased larval mor- tality9, but also increase in growth and faster developmental patterns8,13,14, decreased oxygen consumption rate15 and impaired sensory abilities and behaviour16,17. Further, changed otolith and bone development14,18,19 as well as tissue and developmental damages14,20,21 have been observed. Yet, it has so far been difcult to link all these efects to ftness consequences. 1GEOMAR Helmholtz Centre for Ocean Research Kiel, Marine Evolutionary Ecology, Düsternbrooker Weg 20, 24105, Kiel, Germany. 2Dept. of Economics, Kiel University, Sustainable Fisheries, Wilhelm-Seelig-Platz 1, 24118, Kiel, Germany. 3Nofma AS, Muninbakken 9, NO-9019, Tromsø, Norway. 4Institute for Marine Research, Framsenteret, Hjalmar Johansens gate 14, NO-9007, Tromsø, Norway. 5Centre for Ecological and Evolutionary Synthesis (CEES), Department of Biosciences, University of Oslo, Postboks 1066, NO-0316, Oslo, Norway. *email: fmittermayer@ geomar.de SCIENTIFIC REPORTS | (2019) 9:16908 | https://doi.org/10.1038/s41598-019-52628-1 1 www.nature.com/scientificreports/ www.nature.com/scientificreports Some regulatory mechanisms and capabilities of adult fsh to achieve and sustain acid-base balance such as bicarbonate regulation in the blood plasma or hydrogen and bicarbonate excretion have been identifed22–24, but very little is known about acid-base regulation in early life stages. Gill epithelia, the most important tissue for acid-base regulation in adult fsh22, are only developing in larvae, with a complete absence at hatch in cod larvae (for cod see25). During early development, all proton excretion occurs via chloride cells (Na+/K+ ATPase-rich ionocytes). Tese are located in the skin26 and the primordial gill cavity10. Te detrimental efects of OA on larvae are most likely due to the organism’s limited ability to regulate their acid-base balance or the cost of increased regulation. Te increased cost of acid-base regulation in an acidifed ocean could have potential impacts on the ftness of the larvae14,24,27. With the increasing availability of next generation sequencing techniques, global transcriptome profling in non-model organisms has become possible and afordable. Tis allows to elucidate the molecular basis for observed physiological impairments, and thus, a more detailed characterization of the stress phenotypes28, with the ultimate goal to understand how exactly OA perturbs larval physiology. Tere are, so far, few attempts on untangling the full transcriptomic response to ocean acidifcation stress in fsh. Most studies are based on candi- date gene approaches concentrating on acid-base regulation29,30 in adult fsh. More recent studies have, however, employed whole transcriptome sequencing methods such as RNAseq (for a review see31) and identifed difer- entially expressed (DE) genes. In these studies the focus was mainly on the cellular stress response32 and neuro- logical signalling in adult fsh33. Underlying transcriptomic mechanisms of above mentioned efects during early development remain unclear, and thus, even fewer studies have focused on the gene expression changes to ocean acidifcation in eggs and larvae13. Ocean acidifcation, including a decrease in pH, as well as changes in temperature and hypoxia, can impact the function of many cellular processes and cause damages to cellular macromolecules such as DNA and proteins34. In order to avert damage, cells can react to harmful changes in their cellular environment by means of the cellu- lar stress response (CSR) defned by Kültz34. Te CSR is manifested as the minimal stress proteome, consisting of circa 300 proteins with a core number of highly conserved genes throughout all organisms35. Proteins of the CSR are mainly involved in molecular chaperoning, DNA repair, protein folding, redox regulation and energy metabolism. Cells initiate a CSR in response to damaged macromolecules by increasing transcription of the genes encoding for the stress proteome. Tis leads to increased concentration of CSR proteins while their activity is 35 further controlled by post translational modifcations . Increased pCO2 levels are known to induce diferential expression of genes of the CSR in tissues of adult fsh32 but again very little is known about the CSR in larval fsh under the stress that is imposed by increased pCO2. In order to elucidate the physiological basis of ocean acidifcation efects in Atlantic cod (Gadus morhua) larvae, a key species of ecological and economic importance in the North Atlantic, in response to simu- lated ocean acidifcation levels, this study employed global gene expression profling via mRNA sequencing (RNA-Seq). Atlantic cod embryos and subsequent larvae were exposed to either ambient/control (503 µatm) or end-of-century pCO2 concentrations (1197 µatm). Te sampling dates for the RNA-Seq were based on important physiological changes occurring in the cod larvae36,37 and on the observation of phenotypic traits and mortality9,14 8,13 at those time points. Te phenotypic efects of that experiment have been reported before . Te pCO2 treatment concentrations refecting future ocean states were chosen according to RCP 8.512 and represent the global mean ocean acidifcation level predicted for 2100. Note that on a regional scale (the Barents Sea) these levels will most likely be reached earlier or under more optimistic climate scenarios38. A more detailed description of the experi- mental set-up has been published elsewhere8. In short, end-of-century prediction OA levels were shown to induce dramatic consequences for ftness related measurements, i.e. a doubling in the daily mortality rate (from 7% in 9 ambient to 13% in high pCO2 treatment) and other severe phenotypic diferences like changes in ossifcation rates in the vertebrae, gill development and tissue histology such as liver vacuolization14. Although a small number of studies have examined gene expression in fsh larvae in response to ocean acidi- fcation, these studies have been limited to candidate genes related to stress response39 or acid-base regulation13, which did not allow for the identifcation of potentially important genes and pathways under diferential expres- sion when exposed to simulated ocean acidifcation. To our knowledge, this is the frst study that addresses the efect of OA on the whole transcriptome of marine fsh larvae comparing diferent developmental stages. In addi- tion, special emphasis was placed on gene families connected to the CSR to address the question if Atlantic cod during early developmental stages show signs of cellular stress in response to ocean acidifcation. Results Larval dry weight. To assess the impact of simulated ocean acidifcation levels on growth performance in the Atlantic cod larvae, dry weight measurements were recorded at 5, 15 and 36 days-post-hatch (dph). Te frst two early larval stages measured: 5 dph (Fig. 1a) and 15 dph (Fig. 1b), showed no diferences in dry weight between ambient and high pCO2 treatment (F1,3.73 = 6.00, p = 0.08 and F1,3 = 2.06, p = 0.25, respectively) (full results SI Table 1). A signifcant diference in dry weight was observed at 36 dph (Fig. 1c), where larvae from the 14 high pCO2 treatment were signifcantly heavier than larvae from the ambient treatment . With increasing larval age the variance in dry weight of larvae from the high pCO2 treatment increased compared to the variance in the ambient treatment.